Diagnosis of internal cracks in concrete using vibro-acoustic modulation and machine learning

This paper investigates the utility of physics-informed machine learning models for vibro-acoustic modulation (VAM)–based damage localization in concrete structures. Vibro-acoustic modulation is a nonlinear dynamics-based non-destructive testing method, which was initially developed to perform damage detection and later extended to accomplish damage localization. The VAM-based damage (hidden crack) diagnosis is performed by analyzing the damage index pattern on the surface of the component to arrive at the size and location of the hidden damage. Past investigations have employed heuristically selected damage index thresholds as well as computationally expensive Bayesian estimation methods for VAM-based damage localization in two (surface) dimensions. Compared to these studies, the proposed methodology automates the threshold selection (algorithmic instead of heuristic), increases the speed of the probabilistic damage diagnosis process, and enables the estimation of damage depth. We generate training data (damage index) for the machine learning models using the pertinent nonlinear dynamics (finite element) models using different combinations of test parameters. The (supervised) machine learning models are thus informed by computational physics models. These include two types of artificial neural network (ANN) models: classification models that identify whether a sensor location is damaged or not and regression models that enable Bayesian estimation to obtain the posterior probability distribution of damage location and size. The accuracy of machine learning-based diagnosis is evaluated using both numerical and laboratory experiments. The proposed physics-informed machine learning models for VAM-based damage diagnosis are able to achieve an accuracy of about 60–64% in the validation experiments, indicating the potential of these methods for internal crack detection. The results show that for complex (nonlinear dynamics-driven) diagnostic methods, damage index patterns learned from physics models could be successfully used for damage detection as well as localization.

Structural Health Monitoring in Composite Structures: A Comprehensive Review

This study presents a comprehensive review of the history of research and development of different damage-detection methods in the realm of composite structures. Different fields of engineering, such as mechanical, architectural, civil, and aerospace engineering, benefit excellent mechanical properties of composite materials. Due to their heterogeneous nature, composite materials can suffer from several complex nonlinear damage modes, including impact damage, delamination, matrix crack, fiber breakage, and voids. Therefore, early damage detection of composite structures can help avoid catastrophic events and tragic consequences, such as airplane crashes, further demanding the development of robust structural health monitoring (SHM) algorithms. This study first reviews different non-destructive damage testing techniques, then investigates vibration-based damage-detection methods along with their respective pros and cons, and concludes with a thorough discussion of a nonlinear hybrid method termed the Vibro-Acoustic Modulation technique. Advanced signal processing, machine learning, and deep learning have been widely employed for solving damage-detection problems of composite structures. Therefore, all of these methods have been fully studied. Considering the wide use of a new generation of smart composites in different applications, a section is dedicated to these materials. At the end of this paper, some final remarks and suggestions for future work are presented.

Speakers exhibit a multimodal Lombard effect in noise

In everyday conversation, we are often challenged with communicating in non-ideal settings, such as in noise. Increased speech intensity and larger mouth movements are used to overcome noise in constrained settings (the Lombard effect). How we adapt to noise in face-to-face interaction, the natural environment of human language use, where manual gestures are ubiquitous, is currently unknown. We asked Dutch adults to wear headphones with varying levels of multi-talker babble while attempting to communicate action verbs to one another. Using quantitative motion capture and acoustic analyses, we found that (1) noise is associated with increased speech intensity and enhanced gesture kinematics and mouth movements, and (2) acoustic modulation only occurs when gestures are not present, while kinematic modulation occurs regardless of co-occurring speech. Thus, in face-to-face encounters the Lombard effect is not constrained to speech but is a multimodal phenomenon where the visual channel carries most of the communicative burden.